You are on page 1of 7

The Urban Church and Cancer Control:

a Source of Social Influence in Minority Communities

All the authors are with the Charles R. Drew University of
Medicine and Science, Los Angeles, CA. Dr. Davis is Program
Director, Behavioral Science Research, Drew-Meharry-Morehouse
Consortium Cancer Center (DMMCCC), and Assistant Professor of
Psychiatry and Human Behavior. Dr. Bustamante is an Assistant
Professor of Psychiatry. Dr. Brown is Director of Research and
Assistant Professor of Obstetrics and Gynecology. Dr. WoldeTsadik is Director, Epidemiology-Statistical Core Resource,
DMMCCC, and Associate Professor of Internal Medicine. Dr.
Savage is Associate Director, DMMCCC, and Professor of
Obstetrics and Gynecology. Dr. Cheng is the Project Data
Manager. Ms. Howland is Project Coordinator.
This research was supported by grant No. CA45847 from the
National Cancer Institute, Public Health Service.
Tearsheet requests to Donna T. Davis, PhD, Charles R. Drew
University of Medicine and Science, Department of Psychiatry and
Human Behavior, 1621 E. 120th St., Los Angeles, CA 90059; tel.
310-668-4806; FAX 213-755-1036.



A study was conducted to examine the efficacy of a

church-based model of social influence in improving
access to and participation of underserved minority
women in a cervical cancer control program. The
model expanded on strategies used in previous
hypertension control and health promotion research.
A total of 24 churches, stratified by faith tradition,
randomly selected to participate in the cancer
control program from a pool of 63 churches in a
defined geographic area of Los Angeles County, CA.
Female parishioners ages 21 years and older were
eligible to participate in cervical cancer education
sessions, and screening was offered to adult women
who had not had Papanicolaou tests within the last 2

Church participation rate was 96 percent. Thirty

lay health leaders were selected by the clergy to
serve as messengers, recruiters, and organizers for
their respective congregations. Ninety-seven percent
of these lay health leaders participated in two
training sessions designed to prepare them for their
leadership role. Social support structures such as
child care, meals, or transportation for targeted
women were organized by lay health leaders in 78
percent of the churches.
A total of 1,012 women between the ages of 21 and
89 years attended educational sessions. Forty-four
percent of the eligible women were targeted for
screening because they had not had a Papanicolaou
test within the last 2 years or had never been
screened. Black women were 6.6 times more likely
than Hispanics to have been screened in the past 2
years. Hispanic women were 4.2 times more likely
than African Americans never to have had a
Papanicolaou test or been tested in 3 or more years.
Overall, 90 percent of the women targeted for
screening recruitment presented for tests.

Fifty-two percent of the churches initiated cancer

control activities by the end of the 2-year period
following the culmination of the intervention

The findings suggest that a church-based model of

social influence can leverage the participation of
minority women in cervical cancer control, provide
access to underserved Hispanic women in particular,
and sustain cancer control activities beyond the life
of an intervention program.
The findings further suggest that a more discrete
assessment of screening history may improve the
participation levels of African American women, and
that the gratis offering of screening services may
adversely affect their participation rates.


THE IMPORTANCE OF CANCER as a major health

problem in the United States is well recognized.
Cancer ranks second only to cardiovascular disease as
500 Public Health Reports

a leading cause of death. When black-white mortality

rates are compared, however, African Americans
show an excess cancer mortality of 24 percent (1).

The disproportionate burden of cancer borne by

African Americans is reflected further in morbidity
and survival rates that position black populations, in
comparison with whites, as more likely to develop
cancer and less likely to survive it (1).
Although further research into the mechanisms
underlying these differences and their significance is
warranted; it is known that the improved rates among
whites is attributable largely to their participation in
risk reduction programs (2). In contrast, enlisting the
participation of minority populations in cancer control
has been less effective (1).
In hypertension control programs, approaches to
outreach that produced significant improvements in
participation levels have been identified (3-7). These
approaches can be characterized by three primary
factors: (a) the use of the church as an intermediary,
(b) the inclusion of members of the church social
network as agents of intervention, and (c) the grafting
of program activities onto the natural support systems
of the church community.
Previous research in health promotion has also
concluded that the participation of key members of
the church social network can have a positive
influence on the process and outcome of churchbased interventions (3,5-10). Not much light has been
shed on the process for selecting key members,
however, in previous research. The selection of these
persons has consisted primarily of identifying a
willing party. This tactic can produce extreme
variation in the outcomes of intervention and in study
replication efforts.
In contrast, social influence research by King (11)
identified three attributes of human interaction that
can be used to predict a person's potential to exert
influence over a defined setting of which that person
is a part. These attributes were self-projection
(projecting a positive self-image), interpersonal comfort (demonstrating ease and competence in interpersonal communications with the group), and affiliative
links (being an accepted and responsive member of
the group).
Cancer control models have not used these
strategies or fully applied the concepts of social
influence (12). The results of previous studies on
hypertension control and health promotion programs
suggest that more effective use of indigenous
institutional and social supports could increase the
participation of minority groups in cancer control.
A study was conducted to examine the efficacy of
a church-based model of social influence in reaching
underserved African American and Latina women for
participation in a cervical cancer control program.
The model expanded on strategies used in previous

hypertension control and health promotion research.

The study model included the following overall
* creating the conditions for church-based cancer
* establishing network leadership and social supports,
* implementing the interventions, and
* promoting continuity of leadership initiatives.

Churches. Twenty-four churches, stratified by faith
tradition, were randomly selected from a pool of 63
churches in a defined geographic area of south and
south central Los Angeles, CA. These churches had
responded to a letter offering an opportunity to
participate in Drew University's cervical cancer
control program. Of the responding churches, 36 were
Protestant and 27 were Catholic. The stratification
resulted in equal numbers of Protestant and Catholic
churches participating over a 2-year period. The
church selection procedure did not control for
ethnicity, since the church, not members of the
congregation, was the unit of selection. There was a
close association, however, between ethnicity and
faith tradition. Fifty-four percent of the churches (12
Protestant and 1 Catholic) had a majority congregation of African Americans, and 46 percent of the
churches (11 Catholic) had a predominantly Hispanic
membership. The reported active membership per
church ranged from 50 to 250 people.

Eligible church participants. All women ages 21

years and older were eligible to participate in cervical
cancer education sessions, and screening was offered
to adult women who had not been screened within the
last 2 years. Church registries showed an average
active adult female membership of 60 percent. Approximately 30 to 150 adult women were potential
participants per church, or from 720 to 3,600 for the
combined church population.
Instrumentation. An ethnographic assessment protocol developed and pretested by Davis (13) was used
to guide the selection of leaders within the church
social network. The protocol is an adaptation of a
sociogram methodology developed by King (11) for
the selection of school-based youth leaders. The
ethnographic assessment protocol required three steps.
First, the pastor was asked to identify three female
parishioners who were "natural help givers," or
persons to whom others in the congregation usually
turned for advice. Next, these names were listed on
July-August 1994, Vol. 109, No. 4 501

the scale and the pastor was asked to identify which

of the three persons was best characterized by each of
18 items describing self-projection, interpersonal
comfort, and affiliative links. The person listed most
frequently was named the lay health leader. The
pastor was given an opportunity to confirm this
choice and determine if the first and second place
scorers should serve as a team.

Procedures. The following procedures were used for

each phase of the research process:

Phase 1. Creating the conditions for church-based

cancer control. Creating the conditions for churchbased collaboration in cancer control involved two
major tasks-(a) securing pastoral commitment to the
church role as a partner-in-cancer-prevention and (b)
selecting a lay health leader from the congregation.
The churches that responded to the written contact
also returned a 1-page questionnaire that identified a
contact person (for example, an assistant pastor,
secretary, wife) who could usually be reached at the
church to assist with further communications and
noted the size of the active versus the registered
membership of the church. The pastor and contact
person were called at each church to establish the
first meeting with the pastor.
Very often, pastors asked that the first meeting be
held with the contact person. This approach seemed
to suit better the working style and time constraints
of many clergy. In no instance, however, did the
meeting with the contact person substitute for the
meeting with the pastor. When this alternative
meeting arrangement was suggested, we indicated
that we would not engage in any action with the
church without the pastor's informed consent, and we
explained the extreme importance in having the
leader of the congregation personally endorse the
proposed activities. In those instances, the first
meeting was held with the contact person who would
agree to arrange our meeting date with the pastor
within 2 weeks.
502 Public Health Reports

The meeting dates were arranged for the convenience of pastors and included weekdays, evenings,
and weekends. The meeting times ranged from 1 to
1 /2 hours.
A description of cancer facts that highlighted the
problem of cervical cancer in minority communities
was presented to the pastor. It served as the basis of
our request for partnership with the church. In
addition, a 1-page program overview was prepared
for this meeting that included a listing of proposed
activities and identified supports needed from the
church (for example, pastors' leadership and active
support of the program, 5 minutes during services to
be introduced to the congregation, selection of a lay
health leader among the parishioners, a large meeting
room to conduct education sessions, a secluded room
with doors to conduct the screening, and so forth).
Pastors were asked to sign a declaration indicating
their support as partners in cancer control. The
ethnographic assessment protocol was used to assist
pastors in selecting lay health leaders from the church
social network.

Phase 2. Establishing network leadership and

social supports. Preparing lay health leaders for their
leadership role involved training and the selection of
support structures that would be organized by the
leadership. The lay health leaders were informed in
writing and by telephone of their selection by the
pastor. The pastor also announced his selection
during church services in order to validate their
leadership role. A meeting with each lay health leader
(or team, as applicable) was held to acquaint her with
the problem of cervical cancer, describe her messenger and support roles in the program, and identify
the program staff person who would provide necessary resources (for example, publicity flyers, sign-up
sheets, pamphlets, and brochures) in support of her
role. They were also asked to select two dates from a
list of 10 possibilities for training.
Two groups of lay health leaders were trained to
serve as health advisers and site coordinators for their
respective congregations in two training sessions.
Sessions were limited to 30-45 minutes each in order
to maximize attendance, and refreshments were
served. The training sessions were highly focused and
designed to prepare lay health leaders to (a) deliver
specific messages on screening requirements for early
detection of cervical cancer, the importance of
followup care for persons with abnormal results and
for treatment efficacy and (b) recruit congregation
members and deliver messages with confidence. A
training session was devoted to each objective. A
leadership handbook was prepared for each leader

and included three brochures and five tip sheets for

each of the training areas. Brochures and tip sheets
presented clear and concise information written at a
6th grade reading level.
Each aspect of the program was reviewed in
training sessions. Techniques included a 15-minute
slide presentation (speakers kit) developed by the
National Cancer Institute's Office of Cancer Communications, review, and discussion of the leadership
handbook, and role-play of recruitment and message
delivery strategies. Pre- and post-assessment was not
used to reduce the potential for participant anxiety
when learning new information for which they will
be tested. A training evaluation was administered.
Booster sessions were held as needed, since the
program was implemented during a 2-year period.
The lay health leaders' role during the project
period would include coordinating all dates and times
for program activities at the church and leading the
recruitment of women to education and screening
sessions. In addition, the leaders were asked to select
social support structures that they felt would enable
the participation of women in education and screening sessions, and for which they could draw the
support of the congregation. Lay health leaders
(trainees) also arranged for child care, transportation,
and meals for those attending the sessions.
Phase 3. Implementing the interventions. The
congregation was provided an opportunity to become
acquainted with the program staff members prior to
the implementation of interventions. Program staff
members were introduced by the pastor to the
congregation during services held at least 1 week
before the intervention program. The ethnicities of
the program staff were similar to that of the
congregation. A 5-minute presentation on the program was delivered to the congregation by the
program director. All intervention staff members
remained for the duration of the service in order to
meet and greet parishioners.
Two components constituted the intervention
program-education and screening. All adult women
were recruited to participate in education sessions by
lay health leaders who used church bulletins, service
announcements, publicity flyers, and word of mouth.
A 1-hour session was held immediately following the
Sunday service(s) of the church in a large room,
often the sanctuary. A 20-30 minute survey (English
and Spanish versions) was administered to women in
attendance to determine their screening history. The
speakers' kit (15-minute slide presentation) on cancer
prevention and control was used in the education
sessions. A Spanish version of the English script was

Table 1. Age distribution for 943 women participants in

church-based cervical cancer education project in Los
Angeles, by ethnicity
21-39 years

40-59 years

60-89 years







Hispanic ....
Other .......







Totals ...








Table 2. Years since last Pap test for 943 women participants
in Los Angeles church cancer education project, by ethnicity

3 or more

1 to 2





Hispanic ....
Other .......


















used as appropriate. A 10-minute presentation on the

use of the Papanicolaou (Pap) smear in detecting
cervical cancer and the nature and success of
treatments for abnormal results followed. A question
and answer period was conducted during the last 15
minutes, and women were informed of the screening
A conservative criterion for screening eligibility
was used to identify underserved women within the
church population. Adult women who had not had a
Pap test within the last 2 years were recruited for
screening by lay health leaders. A sign-up list was
added to the screening recruitment strategies
employed for education sessions. Screening also was
conducted immediately following each service or on
one date per church, and performed by two nurse
practitioners. Intake was conducted by program staff
members in a prepared area, and each person's last
screening date was verified. Screening equipment,
including gurneys and screens, were transported to
the church site by program staff members. Two
rooms with doors were used for screening. Women
were informed of the date that results would be
mailed to them and that followup care would be
arranged for those with abnormal results.
Phase 4. Promoting continuity of leadership
initiatives. The pastor and lay health leaders were
involved in the planning for continued promotional
July-August 1994, Vol. 109, No. 4 503

Table 3. Targeted women versus additional women screened

in Los Angeles church cancer education project, by ethnicity
Targeted women

No shows

Number Percent Number Percent



Additional Total
screens screened







Totals .... 372








Other .......

activities. A meeting with the pastor and lay health

leader of each church was held to review participation levels and outcomes of the screening. The
meeting was also used to develop long-term plans for
continued promotion of cancer control messages and
activities. Specific suggestions prepared by program
staff members included annual meetings of the
women auxiliaries on women's health, the use of
project staff members to give presentations or
educational materials, or both, and periodic health
fairs for which program staff members could assist in
identifying clinical resources and personnel.

These are the results for each phase of the
Phase 1. Creating the conditions for church-based
cancer control. Twenty-three of the 24 pastors
consented to participate in the cervical cancer control
program and signed the declaration of partnership, for
a church participation rate of 96 percent. The pastor
who declined was an interim minister and was not the
pastor who responded to our initial contact. The
interim pastor preferred to delay participation in the
program until the permanent pastor had been selected.
The result was 11 Protestant and 12 Catholic
churches participated in the program.
Thirty lay health leaders were selected by the
clergy, one leader each at 16 churches, teams of two
each at 7 churches (4 Protestant, 3 Catholic). Fifteen
percent more of the Catholic churches (four) in
comparison to Protestant churches (two) selected
members of the clergy (for example, nuns, assistant

Phase 2. Establishing network leadership and
social supports. All 30 lay health leaders accepted
their appointed role, and 29 (97 percent) attended
both training sessions. One member of a team of two
504 Public Health Rcports

indicated that she would be supportive but could not

attend training meetings. All trainees evaluated the
training sessions as good to excellent in preparing
them for their roles.
The nature and organization of support structures
varied by church. Eighteen churches (78 percent)
organized support structures. Volunteers at 2 of the
18 churches provided child care, buses, and lunch for
families attending the education and screening sessions held on a Saturday. Twelve of the 18 churches
offered snacks and organized child care. Four
churches provided lunch during which time children
were supervised. Transportation assistance beyond
that normally provided by the church on Sunday
mornings was not needed for these churches, since
interventions were conducted immediately after Sunday services.
Five churches (22 percent) did not formally
organize support structures, but lay health leaders
took personal responsibility for the care of children
as needed during the intervention activities.

Phase 3. Implementing the interventions. Education

and screening were conducted at 23 churches. Lay
health leaders coordinated the dates, times, and
locations for each church. A total of 1,012 women
attended education sessions, and findings are reported
for 943 women (93 percent) who completed the survey. The ethnic distribution of education participants
was 62 percent African American, 35 percent Latino,
and 1 percent Asian, 1 percent Native American, and
1 percent white. Women between the ages of 21 and
39 years represented 53 percent of the sample and
were the largest participating group for blacks and
Latinas (table 1).
The majority of women were married or living
with a mate (51 percent), 30 percent were widowed,
separated, or divorced, and 19 percent were never
married. The great majority of African American
women (95 percent) spent the first 12 years of life in
the United States, whereas 91 percent of the Hispanic
women spent these years in Mexico. (Questions concerning birthplace were not posed so as to avoid
issues of immigration status).
The majority of women were employed (66
percent), and the average annual income range for
African Americans was $20,000 - $24,999 per year,
and $7,500 - $9,999 for Hispanics. The average
number of children ages 17 and younger per household was two for African American women and three
for Hispanics. Thirty percent of the women had
combined household incomes below poverty level.
Seventy-seven percent of the Hispanic women had
less than 12 years of education, whereas 53 percent

of the African Americans had 2 to 3 years of college.

One-third of the women were without health insurance, 54 percent had private coverage, and 13 percent
were covered by Medicare or Medi-Cal. The usual
source of health care was private providers for 59
percent of African American women and 35 percent
of Hispanic women. Emergency rooms and county
clinics were the usual providers of care to the
remaining women.
The average number of women attending education
sessions and completing the survey was 41 per
church. Forty-four percent of the women were
defined as underserved, since they had not had a Pap
test within the last 2 years or had never been
screened (table 2).
Black women were 6.6 times more likely than
Hispanics to have been screened in the past 2 years.
Conversely, Hispanic women were 4.2 times more
likely than African Americans not to have ever
received a Pap test or not to have received a Pap test
in 3 or more years.
Women who had not been surveyed but met the
eligibility requirements were also screened. Of these,
413 were targeted for screening outreach, 490 were
recruited, for an average of 21 women per church.
Ninety-four Hispanic women who had not attended
education sessions presented for screening and 6
Latinas originally targeted for screening did not, for a
total Hispanic group of 374 (table 3). Seventy-two
percent of the African American women targeted for
screening were recruited for Pap tests, and 33 were
no shows. Twenty-three additional Asians, whites,
and Native Americans were screened, and one was
lost to recruitment. Overall, 90 percent (372) of the
women originally targeted for recruitment presented
for screening, and an additional 19 percent (N=117)
were screened.
The majority of women screened were between the
ages of 21 and 39 years, and 37 percent were ages 40
to 59 years (table 4). Hispanic, Asian, Native
American, and white women were heavily represented
among the younger age group, whereas black women
were evenly distributed between the two.
Forty-eight abnormal Pap smears were identified,
for an abnormal rate of 10 percent.
Phase 4. Promoting continuity of leadership
initiatives. Fifty- two percent of the churches (8
Protestant, 4 Catholic) continued the cancer prevention campaign in the first 2 years following the end
of the program period. Table 5 is an overview of the
type and number of the promotional activities.
Program staff members served as resource persons to

lay health leaders. Additional participating agencies

Table 4. Ethnicity and age of 490 women screened in Los

Angeles church cancer education project
21 to 39 years


40 to 59 years

60 to 89 years







Other .......







Totals ...










Table 5. Number of church-initiated cancer control activities

in 12 Los Angeles churches 2 years after intervention
Post-program year

Year 1

Health fairs .3
Interdenominational ministerial
Young adult seminars ........
Health ministry meeting.
Adult Sunday school conference

Year 2




in the church-initiated efforts were the Drew University campus of the Drew-Meharry-Morehouse Consortium Cancer Center, Central Los Angeles Unit of the
American Cancer Society, and Association of Black
Women Physicians.

The high consent rate of pastors, participation rate
of lay health leaders, and recruitment rate of targeted
women add strength to the proposition that social
influence models that use indigenous sources of
social support can exert positive influence on the
participation of minority women in cancer control.
Similarly, the number and variety of cancer control
activities initiated by lay health leaders following the
conclusion of the intervention program offer support
to the social influence approach to institutionalizing
the promotion of cancer control in community-based
The screening profiles of women also present
opportunities for improvements in approaches to
cancer control. The large majority of Hispanic
women reporting screening intervals of 3 years or
more suggests that church-based models may be
particularly valuable in providing access to underserved Hispanic women. The observed pattern of
screening may be due largely to the recent migrant
status of the majority of Latinos in Los Angeles.
July-August 194, Vol. 109, No. 4 505



participation among the 32 percent of targeted black

women who did not show for screening. The potential
impact a token fee may have had on the presenting
women, however, is difficult to evaluate. Since a goal
of cancer control is to maximize the number of
women screened, our experience suggests that caution
should be exercised in the presentation of preventive
services as free. It may be prudent to make this
decision on a church by church basis and to enlist the
assistance of the lay health leader in the decision.

Identifying access opportunities is significant, since
migrants are more likely to use public health facilities
than private ones where Pap tests are usually
conducted in response to a presenting health problem
as opposed to a prevention protocol.
The shorter intervals between screenings reported
by the majority of African American women seem to
reflect more frequent use of preventive care services.
Issues discussed during the question and answer
period following the educational sessions suggest that
African American women make more frequent provider visits, but their receipt of Pap tests may be less
frequent than reported. The majority of these women
indicated that they had assumed a Pap smear was
taken every time they had a pelvic examination, but
they had no knowledge of their Pap smear result, the
correct purpose of the Pap test, or memory of
discussing a Pap test with their provider during the
gynecological visit.
Since survey forms were administered at the
beginning of the educational session and women were
still uncertain by the end of the session, their survey
responses were not revised. The frequency with
which this issue was raised seems to suggest a need
to have assessed, more discretely, the occurrence of
the Pap test versus the pelvic examination. Distinguishing the two may improve needs assessments
conducted by cancer control programs and providers
of preventive gynecological services.
An additional factor influencing the screening
participation rates of African American women may
have been their perception of free services. In the
question and answer period, women tended to
characterize the gratis offering of the Pap test as a
favor to the poor. These women objected to being
perceived as poor, despite their feelings that they
could not afford the cost of regular medical attention.
Women also intimated that "free" services often
results in "substandard" services in black communities. A token fee may have achieved greater
506 Public Health Reports

1. Department of Health and Human Services: Report of the

Secretary's Task Force on Black and Minority Health. Vol.
II, Crosscutting issues in minority health. U.S. Government
Printing Office, Washington, DC, August 1985.
2. Office of Cancer Communications, National Cancer Institute:
Cancer prevention awareness survey. WAVE II, Management
summary. DHHS Publication No. (NIH) 89-2908, National
Institutes of Health, Bethesda, MD, 1986.
3. Eng, E., Hatch, J., and Callan, A.: Institutionalizing social
support through the church and into the community. Health
Educ Q 12: 81-92 (1985).
4. Hatch, J.: Outreach in Chatman County. N C Med J 848:
633-635 (1987).
5. Hatch, J., and Lovelace, K.: Involving the southern rural
church and students of health professions in health education.
Public Health Rep 95: 23-25, January-February 1980.
6. Levin, J.: The role of the black church in community
medicine. J Natl Med Assoc 76: 477-483, May 1984.
7. Hatch, J., Cunningham, A., Woods, W., and Snipes, F.: The
fitness through churches project: description of a communitybased cardiovascular health promotion intervention. Hygiene
V: 9-12 (1986).
8. Mollica, R., Streets, F., Boscarino, J., and Redlick, F.: A
community study of formal pastoral counseling activities of
the clergy. Am J Psychiatry 143: 323-328 (1986).
9. Hatch, J.: The North Carolina Baptist church program. Urban
Health 10: 70-71, May 1981.
10. Hatch, J.: The rural minister's role in health promotion. J
Religious Thought 46: 69-77, summer-fall 1989.
11. King, L. M.: Leadership training model. Fanon Center
Monograph, Charles R. Drew University of Medicine and
Science, Los Angeles, CA, 1981.
12. Lacey, L., et al.: An urban community-based cancer
prevention screening and health education intervention in
Chicago. Public Health Rep 104: 536-541, NovemberDecember 1989.
13. Davis, D. T.: Partners in prevention program: evaluation
report. Office of Cancer Communications, National Cancer
Institute, Bethesda, MD, 1990.